The DHS Program User Forum: Tanzania

Home » Countries » Tanzania » Use of Clusters

Show: Today's Messages :: Show Polls :: Message Navigator

Use of Clusters [message #11489]

Thu, 29 December 2016 13:33

Robert
Messages: 19
Registered: April 2015
Location: Dar Es Salaam, Tanzania

Member

Dear all, I want to include some variables like contraceptives use and prenatal care using clusters level. How do I go about it?
Thank you so much

R.Lihawa

Report message to a moderator

Re: Use of Clusters [message #11505 is a reply to message #11489]

Tue, 03 January 2017 19:01

Reduced-For(u)m
Messages: 292
Registered: March 2013

Senior Member

If you use Stata, you can just:

egen ClusterX = mean(X), by(cluster)

And then include "ClusterX" as a variable in your regressions.

I'm sure there is very similar code for creating sub-group means in other programs too. And there are some inference (p-value) issues down the road, but if you cluster your regressions at the level of your clustered X, you should be fine. Unless I missed something...

Report message to a moderator

Previous Topic:	How to generate proportion of women reporting to have used contraceptives at regional level
Next Topic:	Enumeration Areas

Goto Forum:

-=] Back to Top [=-

[ Syndicate this forum (XML) ] [

] [

]

Current Time: Thu Nov 28 16:40:41 Coordinated Universal Time 2024